Lockdown Mode: A New Security Layer for ChatGPT

OpenAI announced a new security configuration called Lockdown Mode, designed specifically for individuals who operate in high‑risk digital contexts. The company emphasizes that most users do not need to enable the feature because the standard protections built into ChatGPT already provide sufficient safety. However, for groups such as journalists, activists, and professionals handling sensitive information, the added restrictions aim to reduce exposure to exploitation, manipulation, or unintended data leaks.

Lockdown Mode works by tightening the model’s interaction surface. Web browsing, for example, is limited to cached content only, preventing the system from pulling live data from the internet. While this reduces the chance of transmitting sensitive information, it also means that search results may be outdated or incomplete. Image handling is similarly constrained: ChatGPT can no longer embed images in its replies, though users can still upload their own images and continue to use OpenAI’s image‑generation tools.

Advanced analytical features are also disabled. Deep Research, which enables complex multi‑step analysis, is turned off, as is Agent Mode, which would otherwise allow the model to perform more autonomous actions. Restrictions extend to networking and file operations: code generated on the Canvas platform cannot be approved for network access, and the model cannot download files for analysis. It can still work with files that users manually upload, but it will not fetch any content on its own.

Availability and Administration

Lockdown Mode is currently offered to four specific user groups: ChatGPT Enterprise, Education, Healthcare, and Teachers. OpenAI has indicated that the feature will be rolled out to consumer and team plans in the coming months. For organizations, the mode is managed at the workspace level. Administrators can create a dedicated role for Lockdown Mode and assign it to selected users, allowing tighter controls without altering settings for the entire workspace.

Related Updates

In parallel with the Lockdown Mode rollout, OpenAI introduced a capability that lets users steer Deep Research using content from their own sites and applications, giving more control over how results are gathered. The company also updated ad settings for free accounts, offering an option to avoid ads while noting that usage limits may change as a result.

Overall, Lockdown Mode represents a focused effort by OpenAI to provide a more secure environment for users who face heightened digital threats, while preserving the broader functionality of ChatGPT for the majority of its audience.

Usado: News Factory APP - descubrimiento de noticias y automatización - ChatGPT para Empresas